- Email: [email protected]
A first-choice combined oral contraceptive influences general well-being in healthy women: a double-blind, randomized, placebo-controlled trial
ORIGINAL ARTICLE: CONTRACEPTION
A first-choice combined oral contraceptive influences general well-being in healthy women: a double-blind, randomized, placebo-controlled trial Niklas Zethraeus, Ph.D.,a Anna Dreber, Ph.D.,b Eva Ranehill, Ph.D.,c,d Liselott Blomberg, B.Sc.,e Fernand Labrie, M.D., Ph.D.,f Bo von Schoultz, M.D., Ph.D.,d,e Magnus Johannesson, Ph.D.,b n Hirschberg, M.D., Ph.D.e,g and Angelica Linde a Department of Learning, Informatics, Management and Ethics, Karolinska Institutet, Stockholm, Sweden; b Department of Economics, Stockholm School of Economics, Stockholm, Sweden; c Department of Economics, University of Zurich, Zurich, Switzerland; d Department of Economics, University of Gothenburg, Gothenburg, Sweden;e Department of Obstetrics and Gynecology, Karolinska University Hospital, Stockholm, Sweden; f Laval University, Quebec City, Montreal, Canada; and g Department of Women's and Children's Health, Karolinska Institutet, Stockholm, Sweden
Objective: To determine whether there is a causal effect of oral contraceptive (OC) treatment on general well-being and depressed mood in healthy women. Design: Double-blind, randomized, and placebo-controlled trial. Setting: University hospital. Patient(s): Three hundred and forty healthy women aged 18–35 years randomized to treatment, of whom 332 completed the data collection at follow-up evaluation. Intervention(s): A combined OC (150 mg levonorgestrel and 30 mg ethinylestradiol) or placebo for 3 months of treatment. Main Outcome Measure(s): Primary outcome measures: global score of Psychological General Well-Being Index (PGWBI) and the Beck Depression Inventory (BDI); secondary outcome measures: six separate dimensions of the PGWBI. Result(s): The OC treatment statistically significantly decreased general well-being compared with placebo 4.12 (95% CI, 7.18 to 1.06). Furthermore, OC decreased the following PGWBI dimensions compared with placebo: positive well-being 3.90 (95% CI, 7.78 to 0.01), self-control 6.63 (95% CI, 11.20 to 2.06), and vitality 6.84 (95% CI, 10.80 to 2.88). The effect of OC on depressive symptoms and on the PGWBI dimension depressed mood were not statistically significant. Conclusion(s): This study demonstrates a statistically significant reduction in general well-being by a first-choice OC in comparison with placebo in healthy women. We found no statistically significant effects on depressive symptoms. A reduction in general well-being should be of clinical importance. (Fertil SterilÒ 2017;-:-–-. Ó2017 by American Society for Reproductive Medicine.) Key Words: Depression, oral contraceptives, quality of life, randomized clinical trial, well-being Discuss: You can discuss this article with its authors and with other ASRM members at https://www.fertstertdialog.com/users/ 16110-fertility-and-sterility/posts/15214-23476
Received November 29, 2016; revised January 28, 2017; accepted February 24, 2017. N.Z. has nothing to disclose. A.D. has nothing to disclose. E.R. has nothing to disclose. L.B. has nothing to disclose. F.L. has received personal fees from EndoCeutics Inc. B.V.S. has nothing to disclose. M.J. has nothing to disclose. A.L.H. has nothing to disclose. Supported by research grants from Jan Wallander and Tom Hedelius Foundation (P2010–0133:1, P2012–0002:1, and P2013–0156:1), Knut and Alice Wallenberg Foundation (Wallenberg Academy Fellows grant to A. Dreber), Swedish Council for Working Life and Social Research (2006–1623), the Swiss National Science Foundation (100010–149451), the Swedish Research Council (20324), Karolinska Institutet, and the regional agreement on medical training and clinical research (ALF) between Stockholm County Council and Karolinska Institutet (20130313). The funders of the study had no role in study design, data collection, data analysis, data interpretation, or writing of the report. All authors had full access to all the data in the study and were responsible for the decision to submit for publication. N.Z., A.D., and E.R. should be considered similar in author order. n Hirschberg, M.D., Ph.D., Department of Obstetrics and Gynecology, Karolinska University Hospital, Stockholm SE-171 76, Reprint requests: Angelica Linde Sweden (E-mail: [email protected]). Fertility and Sterility® Vol. -, No. -, - 2017 0015-0282/$36.00 Copyright ©2017 American Society for Reproductive Medicine, Published by Elsevier Inc. http://dx.doi.org/10.1016/j.fertnstert.2017.02.120 VOL. - NO. - / - 2017
1
ORIGINAL ARTICLE: CONTRACEPTION
T
he introduction of the oral contraceptive pill (OC) more than 50 years ago has had a tremendous impact on women's reproductive health and position in society. Today, it is estimated that over 100 million women worldwide are current users (1). Although OC pills are recognized as highly effective and generally safe, there is still remarkably poor adherence to this contraceptive method. In fact, numerous surveys show a high discontinuation rate and irregular use in as many as 60% of women (1–3). Poor adherence and contraceptive failure have been related to certain side effects such as negative mood changes and depressive symptoms (4–6). Some 4% to 10% of OC users report depressive symptoms (7–9). However, very few placebo-controlled studies have addressed this issue, and they have been performed in women with premenstrual symptoms (10), sterilized women (11), women with dysmenorrhea (12), and in those with previous depressive symptoms using OCs (13). Most of these relatively small studies have not found any statistically significant effect of OC treatment on depressive symptoms (10–12). However, one placebocontrolled study in healthy women demonstrated a statistically significant impairment of mood symptoms by OC use (14); another study showed worsening of depressive symptoms in women with previous experience of negative mood symptoms while taking OCs (13). Recently, a large cohort study also demonstrated an association between hormonal contraception and subsequent use of antidepressants (15). The daily intake of OCs and a positive attitude of the user are a necessary prerequisites for efficacious oral contraception. Irregular intake might reflect a negative influence on the individual woman's well-being in a more general sense than specific depressive symptoms. Apart from depressed mood, well-being also implies dimensions like vitality, selfcontrol, and perception of general health. There is very little information about causal effects of OC use on general wellbeing in healthy women. We found only two small placebocontrolled studies, both with inconclusive results (10, 11). We have filled a knowledge gap regarding a possible causal effect of OC treatment on general well-being and depressive symptoms in healthy women. We performed a double-blind, randomized, placebo-controlled trial on a combined OC in 340 young, healthy women. Our hypothesis was that OCs decrease general well-being and increase depressive symptoms (7–9, 13–15).
MATERIALS AND METHODS Study Design The study was a single center, randomized, double-blind, placebo-controlled, and parallel study conducted at the Karolinska University Hospital, Stockholm, Sweden. The study was performed in accordance with good clinical practice and the World Medical Association Declaration of Helsinki's ethical principles for medical research involving human subjects. The protocol and the informed consent form were approved by the local ethics committee in Stockholm (2010/1075-31/ 1) and the Swedish Medical Products Agency (151:2010/ 44213). The trial is registered at the EudraCT registry, registration number EudraCT 2010-020824-23. 2
Study Population The sample comprised women willing to start using OCs. The inclusion criteria were 18 to 35 years of age, body mass index (BMI) 19–30, a regular menstrual cycle (25–33 days), using nonhormonal contraceptive at the study start, and fluency in the Swedish language. The exclusion criteria were smoking, hypertension, risk factors for thromboembolism and cardiovascular disease, severe migraine, diabetes mellitus, liver disease, pancreatitis, history of hormone-dependent cancer, undiagnosed bleeding, and pregnancy. Intake of any sex steroid hormones during the last 6 weeks before beginning the study was not allowed. Those fulfilling all the inclusion criteria and none of the exclusion criteria were included in the study. All women were instructed to use nonhormonal contraceptive methods during the study. Free condoms were distributed.
Randomization Participants were randomly assigned to treatment with a combined monophasic pill (Neovletta; Bayer Schering Pharma), containing 150 mg of levonorgestrel and 30 mg of ethinylestradiol, or placebo during a period of 3 months. Randomization was provided by a contract organization (APL, Pharma Specials). A balanced block randomization [1:1] was performed with 10 participants in each block of fixed size. The study investigators, research coordinators, and the participating women were blinded to the treatment allocation. Participants took one capsule every day for 21 days, followed by a break during 7 days. Each 28-day period defined one treatment cycle.
Procedures During the baseline visit, scheduled for the early follicular phase of the menstrual cycle (cycle days 1–7), a general health control was performed that included weight, height, and blood pressure measurements. Participants also provided data on education and marital status. At baseline, a blood sample (20 mL) was collected for analyses of hormones and binding proteins. After the blood sample collection, participants filled out two surveys on general well-being and depressive symptoms: the Psychological General Well-Being Index (PGWBI) (16) and the Beck Depression Inventory (BDI) (17). The final data collection was performed at the end of the third treatment cycle. Participants underwent the same procedures as at baseline: weight and blood pressure measurements, blood sampling for hormone determination, and assessment of general well-being and depressive symptoms. All the surveys were computerized and were completed by the participants at the study center between 08:00 and 18:00 the same day.
Outcomes Our primary outcome measure of general well-being was the normalized global score of the PGWBI (16), and our secondary outcome measures of general well-being were the normalized scores on each PGWBI dimension. The PGWBI measured VOL. - NO. - / - 2017
Fertility and Sterility® general well-being during the last month and consists of 22 items, where each item has six response choices and the value of each item ranges from 0 to 5. This scale has had satisfactory internal construct validity when tested with modern psychometric techniques (18). The PGWBI consists of the following six dimensions: anxiety (included five items), depressed mood (three items), positive well-being (4 items), self-control (3 items), general health (3 items), and vitality (4 items). The global PGWBI score was computed as the sum of scores on each item (that is, each of the 22 items was given the same weight). The global score of PGWBI varies between 0 (poor quality of life) and 110 (good quality of life). A score of %72 in PGWBI indicates moderate to severe distress in general well-being. The score on each PGWBI dimension (anxiety, depressed mood, positive well-being, self-control, general health, and vitality) was defined as a secondary outcome measure. The total score for the different dimensions was anxiety (0–25), depressed mood (0–15), positive wellbeing (0–20), self-control (0–15), general health (0–15), and vitality (0–20). To facilitate comparison across studies and for comparisons with other quality-of-life instruments, the total score on each dimension and the global score of PGWBI was normalized to 0–100. The global score of the BDI (17) was our primary outcome measure of depressive symptoms. The BDI consists of 21 items, where each item has four response choices, and the value of each item ranges from 0 to 3. The BDI measured the following 21 depressive symptoms and attitudes during the last week: sadness, pessimism/discouragement, sense of failure, dissatisfaction, guilt, expectation of punishment, self-dislike, selfaccusation, suicidal ideation, crying, irritability, social withdrawal, indecisiveness, body-image distortion, work retardation, insomnia, fatigability, anorexia, weight loss, somatic preoccupation, and loss of libido. A global score of BDI is computed as the sum of scores on each item, and ranges from 0–63, where a higher value reflects more severe depressive symptoms/attitudes. A score R20 in the BDI indicates moderate to severe depressive symptoms.
SHBG, and a fixed albumin concentration based on an equation system derived from the law of mass action (20).
Statistical Analysis Before completing the data collection and before breaking the randomization codes we posted a predefined analysis protocol at the Open Science Framework Web site (freely available at https://osf.io/he8nb/) on August 21, 2015. The main analysis was performed as an intention-to-treat analysis (ITT) and included all women who completed the data collection. To test whether there was a statistically significant difference in the average value of an outcome measure between the OC group and the placebo group, we used an independent samples t-test (not assuming equal variances). The tests were based on the difference in the value of the outcome measure between the final data collection and the baseline data collection for each participant, and all statistical tests were two-sided (alpha level, 0.05). We specified two subgroup analyses in the protocol for two subgroups of moderate to severe distress in general well-being and depressive symptoms. These two subgroups were defined as those who had scores of %72 in PGWBI (moderate to severe distress) before normalization of the score and those who had scores R20 in BDI (severity level: moderate to severe). In exploratory analyses, we also estimated the correlation between the pretreatment and posttreatment change in serum levels of total and free testosterone and the change in general well-being and depressive symptoms, using the Pearson and the Spearman correlation coefficients. The sample size of 340 gives us a statistical power of 80% to find a statistically significant difference at the 5% level in general well-being (PGWBI) of 4.3 units measured on the PGWBI normalized scale between 0 and 100, and 2.1 units in depressive symptoms measured on the BDI scale between 0 and 63. In the Supplemental Appendix we provide a complete methods section with details about the study population, study design, treatment, general well-being and depressive symptom measurements, safety assessments, hormone measurements, and statistical analysis.
Safety Assessments Adverse events and severity were assessed at each scheduled contact and whenever there was an event. The women were reminded to use nonhormonal contraception. At each visit, vital signs including blood pressure and weight were recorded, and a pregnancy test was performed.
Hormone Measurements Quantification of testosterone and estradiol in serum was performed by the validated, highly sensitive and specific liquid chromatography with tandem mass spectrometry (LC-MS/ MS) method performed at Endoceutics (Quebec, Canada) (19). Serum concentrations of follicle-stimulating hormone (FSH) and luteinizing hormone (LH) were determined by fluorescence immunoassays (Auto Delfia; Perkin Elmer), and sex-hormonebinding globulin (SHBG) by chemiluminescent enzyme immunometric assay (Roche Diagnostics). Concentrations of free testosterone were calculated from values of total testosterone, VOL. - NO. - / - 2017
RESULTS About 1,000 women, were initially screened by phone, and 347 women were assessed for eligibility at a screening visit. Three hundred and forty women were included in the study between February 3, 2012, and August 24, 2015, and were randomly assigned to either OC or placebo (169 in the OC group, 171 in the placebo group). Seven participants were lost to follow-up evaluation, and one participant did not complete the data collection due to an administrative error (Fig. 1). All the 332 women who completed the data collection at follow-up evaluation were included in our main analysis, including the three women who discontinued treatment but completed the data collection. There were no statistically significant differences in any of the baseline characteristics between the two treatment groups (P>.05) (Table 1). We hypothesized that OCs would reduce general wellbeing and increase depressive symptoms. Consistent with this hypothesis, there was a statistically significant difference in 3
ORIGINAL ARTICLE: CONTRACEPTION
FIGURE 1
340 Underwent randomization
169* Were assigned to and received oral contraceptives (150 μg levonorgestrel and 30 μg ethinylestradiol)
4 Discontinued drug and were lost to follow-up - 1 anxiety - 1 increased appetite - 1 depression - 1 panic disorder 1 Continued drug but did not complete data collection due to an administrative error
2 Discontinued drug but completed data collection† - 1 mood changes - 1 lost the pills
162 Continued drug and completed data collection†
171 Were assigned to and received placebo
3 Discontinued placebo and were lost to follow-up - 1 forgot pills when travelling - 1 depression - 1 unknown
1 Discontinued placebo but completed data collection† - 1 nausea
167 Continued placebo and completed data collection†
Trial profile.* One participant initially randomized to the treatment group discontinued treatment due to an egg donation but was later randomized into the study again. This explains the uneven number of participants in the two groups. yAll participants that completed the data collection at follow-up were included in our main analysis (N¼164 in the oral contraceptive group and N¼168 in the placebo group). Zethraeus. Oral contraceptives and well-being. Fertil Steril 2017.
the change in the global score of PGWBI between the OC and the placebo groups of 4.12 (95% CI, 7.18 to 1.06; P¼ .0085) (Fig. 2A). We also found a statistically significant negative impact of OC on well-being in the hypothesized direction for the following three PGWBI dimensions (Fig. 2B): positive well-being 3.90 (95% CI, 7.78 to 0.01; P¼ .0492), self-control 6.63 (95% CI, 11.20 to 2.06; P¼ .0046), and vitality 6.84 (95% CI, 10.80 to 2.88; P¼ .0008). The effect sizes expressed as Cohen's D (21) were 0.29 for the global PGWBI score, 0.22 for reduced positive well-being, 0.34 for self-control, and 0.41 for the vitality dimension. These effect sizes are considered small to medium. For the other three PGWBI dimensions, anxiety 2.08 (95% CI, 6.15 to 1.99; P¼ .3155), depressed mood 3.67 (95% CI, 7.51 to 0.17; P¼ .0613), and general health 2.15 (95% CI, 5.24 to 0.94; P¼ .1723), we did not find a statistically significant difference between the OC and the placebo group (Fig. 2B). For the BDI, our primary outcome measure of depressive symptoms, the effect was not statistically significant 0.85 (95% CI, 0.66 to 2.36; P¼ .2664) (Fig. 3). 4
The effect size and statistical significance only changed marginally in our two robustness tests when we also included individuals who did not complete the data collection (n ¼ 340) or only included participants who did not discontinue treatment (n ¼ 329) (P¼ .0088 and .0073 for the PGWBI; .0502 and .0489 for PGWBI dimension ‘‘positive well-being’’; .0048 and .0047 for the PGWBI dimension ‘‘self-control’’; and finally .0008 and .0005 for the PGWBI dimension ‘‘vitality’’). Neither the BDI (P¼ .2684 and .2643), nor any of the other PGWBI domains (anxiety: P¼ .3168 and .3189; depressed mood: P¼ .0622 and .0583; and general health: P¼ .1745 and .1110) were statistically significant in any of the two robustness tests. The proportion of women with moderate to severe distress in general well-being (%72 on the PGWBI) and with moderate to severe depressive symptoms (R20 on the BDI) was comparable between the groups before treatment (Table 1). We found that 35% the OC and placebo groups had moderate to severe distress in general well-being at baseline, whereas 7% in the OC and placebo group had moderate to severe VOL. - NO. - / - 2017
Fertility and Sterility®
TABLE 1 Baseline characteristics of the intention-to-treat population. Characteristic
Placebo (n [ 171)
OC (n [ 169)
Age (y) 24.1 4.0 23.3 3.6 Educationa Primary school 5 (2.9) 6 (3.6) Secondary school 53 (31.2) 58 (34.7) University studies 64 (37.6) 56 (33.5) (ongoing) University degree 48 (28.2) 47 (28.1) In a stable relationshipa Yes 100 (58.8) 97 (57.7%) 22.6 2.8 22.5 3.1 BMI (kg/m2) Participants with moderate to severe distress in general well-being %72 in the PGWBI 60 (35.1) 59 (34.9) Participants with moderate to severe depressive symptoms R20 in the BDI 12 (7.0) 12 (7.1) 68.5 14.6 69.0 13.5 Score on PGWBIb b Scores on PGWBI dimensions Anxiety 63.5 18.2 62.9 18.5 Depressed Mood 79.3 16.3 78.8 15.2 Positive Well-Being 60.7 18.0 61.6 17.3 Self-Control 71.1 19.6 71.7 18.9 General Health 84.5 13.6 84.7 12.9 Vitality 60.2 17.1 63.0 15.8 c 7.3 7.3 7.8 6.9 Score on BDI Serum levels of blood sample testsd FSH (IE/L) 5.4 1.7 5.4 1.7 LH (IE/L) 4.9 2.4 4.8 2.2 Total T (pg/mL) 246.4 93.7 244.3 88.7 SGBG (nmol/L) 78.1 46.7 72.6 27.3 Free T (pg/mL) 3.2 1.3 3.2 1.3 Estradiol (pg/mL) 33.5 19.1 34.4 25.1 Note: Data presented as mean standard deviation or n (%). The body mass index (BMI) is the weight in kilograms divided by the square of the height in meters. There were no statistically significant differences between the two treatment groups (P>.05), as assessed with the use of the chi-square test for categorical data and the independent samples t-test (not assuming equal variances) for continuous data. BDI ¼ Beck Depression Inventory; BMI ¼ body mass index; FSH ¼ follicle-stimulating hormone; LH ¼ luteinizing hormone; OC ¼ oral contraceptive; PGWBI ¼ Psychological General Well-Being Index; SHBG ¼ sex hormone-binding globulin; T ¼testosterone. a There were three missing values on the education variable—two in the OC group and one in the placebo group—and two missing values on the marital status variable (one in the OC group and one in the placebo group). b The PGWBI (and each PGWBI dimension) ranges from 0 to 100, with higher values indicating greater general well-being. The PGWBI is defined as the average of the normalized score on each PGWBI item where each item is given the same weight. c The Beck Depression Inventory (BDI) scores range from 0 to 63, with higher scores indicating more severe depressive symptoms. d There were two missing values in the OC group for FSH, LH, SHBG, and free T and one missing value for total T and estradiol in the OC group. Zethraeus. Oral contraceptives and well-being. Fertil Steril 2017.
depressive symptoms. After treatment, 44% in the OC group and 38% in the placebo group had moderate to severe distress in general well-being, whereas 7% in the OC group and 7% in the placebo group had moderate to severe depressive symptoms. There was no statistically significant treatment effect for the primary outcome measures or the secondary outcome measures in these subgroups, but it should be noted that the study was not powered for comparisons within subgroups. As expected, the serum levels of FSH, LH, and total and free testosterone decreased after 3 months of active treatment, and the change in hormone levels was statistically significantly different between the OC and the placebo groups, supporting that the participants were compliant with the study treatment (independent samples t-test [not assuming equal variances], P< .001) (Supplemental Table 1). VOL. - NO. - / - 2017
In exploratory analyses, we tested whether the size of the pretreatment and posttreatment changes in serum levels of total and free testosterone were correlated with the change in general well-being and depressive symptoms. None of these correlations were statistically significant in the OC group (Pearson and Spearman correlations, P>.05, twosided), and only two correlations were statistically significant in the placebo group (Supplemental Tables 2 and 3). We detected adverse events in 21.0% of the women in the OC group and 12.0% in the placebo group (Supplemental Table 4). There were four serious adverse events in total, including two pregnancies and one ovarian cyst in the placebo group and one panic disorder in the OC group. The most common reported adverse events in the OC group were bleeding disturbances (n ¼ 14), anxiety and mood changes (n ¼ 12), acne (n ¼ 5), and appetite changes (n ¼ 3). Anxiety and mood changes (n ¼ 4) were the most common reported adverse events in the placebo group.
DISCUSSION Here we report results from the first large randomized, placebo-controlled trial of a combined OC on general wellbeing and depressive symptoms in healthy women. We found a statistically significant reduction in general well-being in women who used a levonorgestrel-containing OC for 3 months compared with placebo, as measured with the established PGWBI instrument. There were also negative effects on the specific PGWBI dimensions of positive well-being, selfcontrol, and vitality. Thus, the results were consistent with our hypothesis that in some women OC use could cause a decrease in general well-being. The difference in the primary outcome measure PGWBI between the OC and the placebo group corresponds to a 6.0% reduction of the baseline value. The corresponding reductions in the three significant PGWBI dimensions (positive well-being, self-control, and vitality) were 6.4%, 9.3%, and 11.1%, respectively. These small to medium reductions, in terms of Cohen's D terminology (21), in well-being are likely to be of clinical importance for individual women and may be one explanation for the high discontinuation rate. For the other primary outcome measure, depressive symptoms as assessed by the BDI, we found no statistically significant negative effect. Thus, our hypothesis of increased depressive symptoms could not be confirmed. Moreover, the PGWBI dimension of depressed mood was not statistically significantly different from the controls. The nonstatistically-significant effect on depressive symptoms is in line with most previous placebo-controlled studies (10–12). However, higher scores of depressed mood with OC treatment in comparison with placebo have been reported in two previous placebo-controlled studies (13, 14). Furthermore, an association between hormonal contraception and a first diagnosis of depression, especially among adolescents, was recently suggested in a large cohort study (15). The mechanism behind impaired well-being by OC use in healthy women is not known. We found an increase in SHBG and consequently a decrease in total and free testosterone 5
ORIGINAL ARTICLE: CONTRACEPTION
FIGURE 2
Mean changes from baseline to end of follow-up in scores for the Psychological General Well-Being Index (PGWBI). (A) Results for the global score of PGWBI (the average of the score on each item). (B) Results for each PGWBI dimension. The global score of PGWBI and the score on each PGWBI dimension range from 0 to 100, with higher scores indicating increased well-being. Error bars denote 95% confidence intervals. Zethraeus. Oral contraceptives and well-being. Fertil Steril 2017.
accompanying OC treatment. Some studies have demonstrated positive effects of testosterone treatment on general well-being in postmenopausal women (22, 23). Consequently, a reduction in free testosterone by OC use might cause reduced general well-being. However, we found no statistically significant correlations in the OC group between the change in serum levels of total and free testosterone and changes in general well-being or depressive symptoms. Another possible mechanism whereby OCs might affect general well-being is a direct negative effect of the progestin component on the brain. Sex hormones can be classified as neurosteroids because they can be synthesized in the central and peripheral nervous system (24). They bind to their specific nuclear receptors and act via genomic effects but could also influence brain function via more rapid nongenomic mechanisms and indirectly through neurotransmitter systems, including the serotonin and gamma-aminobutyric acid 6
(GABA) systems (25). Estrogen has been shown to enhance brain excitability, whereas progesterone seems to have an opposite and inhibitory influence (26). Progesterone and several progesterone metabolites exert a sedative effect on the brain. Some metabolites (e.g., allopregnanolone) could even be regarded as anesthetic agents (26). Clearly, such influence could affect the sense of vitality and reduce general wellbeing. Cyclic negative mood symptoms such as irritability, loss of impulse control, depression, and fatigue have been related to an increase of progestogenic neurosteroids during the luteal phase of the menstrual cycle (27, 28). Different progestins and combined OCs have been reported to provoke similar symptoms (29). Although we did not find statistically significant effects on depressive symptoms, it is possible that a direct progestin-induced central nervous system effect could be the underlying mechanism behind reduced well-being, self-control, and vitality in the OC group. VOL. - NO. - / - 2017
Fertility and Sterility®
FIGURE 3
Mean changes from baseline to end of follow-up in scores for the Beck Depression Inventory (BDI). The BDI ranges from 0 to 63, with higher scores indicating more depressive symptoms—that is, a positive change reflects an increase in depressive symptoms. Error bars denote 95% confidence intervals. Zethraeus. Oral contraceptives and well-being. Fertil Steril 2017.
Combined OCs are more or less equally effective as contraceptive agents, but they may differ in other aspects such as risk profile, side effects, and compliance. Most combined OCs contain the synthetic ethinylestradiol as an estrogen component, whereas a few recently developed contain estradiol. In contrast, there are several different progestins with varying degrees of androgenic activity combined with the estrogen component. Androgenic progestins in combination with ethinylestradiol (second-generation OCs) seem to be associated with a lower risk of venous thromboembolism and are therefore recommended as the first choice for contraception in some countries (30). In the present study, we used a levonorgestrel-containing OC, which is the drug most commonly prescribed in Sweden. On the other hand, negative side effects such as mood changes, acne, increased appetite, weight gain, and sexuality may be more frequent with androgenic OCs compared with antiandrogenic OCs (third- and fourth-generation OCs) (31-33). Antiandrogenic OCs may even have positive effects on mood. In placebo-controlled trials, combined OCs containing drospirenone (antiandrogenic fourth-generation OC) have demonstrated beneficial effects in women with premenstrual syndrome (34). Also, a shorter hormone-free interval by the 24/4 regimen may contribute to the beneficial effect because adverse mood symptoms are most pronounced during the pillfree interval (9). Thus, our findings may not be valid for all OCs. Additional studies on antiandrogenic OCs are needed. If the results from alternative OCs show less or no adverse effect on general well-being, this may be reason enough to reconsider recommendations of first-choice combined contraception. The strengths of our present study are the large number of participants and the double-blind placebo-controlled design. VOL. - NO. - / - 2017
However, one limitation was the short study period because it is well-known that several adverse effects vanish over time. Still, mood changes usually do not disappear, and a further deterioration is also possible. Furthermore, positive and negative expectations by the women under treatment may have influenced the results. The use of OCs will change the amount of and cyclicity of menstrual bleeding and thus patient blinding in reality may not have been complete. However, it is also known that women who usually have regular cycles can become anovulatory during a test situation. Even if questions on menstrual cyclicity were avoided, women may have suspected that they were given active or inactive treatment. Such bias, however, could work in both directions. Additional strengths of the study are the use of well-established and validated instruments for assessment of general well-being (PGWBI) and depressed mood (BDI), and the highly specific, validated, and accurate method for testosterone analysis.
CONCLUSION Our present study demonstrates a statistically significant reduction in general well-being with the use of a popular levonorgestrel-containing OC for 3 months in comparison with placebo in healthy women. Furthermore, the PGWBI dimensions of positive well-being, self-control, and vitality were reduced. However, we found no statistically significant effects on depressive symptoms in this group of healthy women. Even a modest reduction in general well-being during OC intake should be of clinical importance and may explain the high discontinuation rate and irregular use of this contraceptive method. This possible side-effect thus has to be recognized and discussed during contraceptive counseling. It is vitally important to evaluate alternative OC 7
ORIGINAL ARTICLE: CONTRACEPTION combinations in large placebo-controlled studies to identify those with the most beneficial profile. Acknowledgments: The authors thank Elisabeth Berg and Mesfin Tessma for valuable comments on the analysis protocol; Berit Legerstam, Siv R€ odin Andersson, and Agneta Berge for research assistance; and Johan Almenberg for proofreading the manuscript.
15. 16.
17.
18.
REFERENCES 1. 2. 3.
4.
5.
6.
7.
8.
9. 10.
11.
12. 13.
14.
8
Brynhildsen J. Combined hormonal contraceptives: prescribing patterns, compliance, and benefits versus risks. Ther Adv Drug Saf 2014;5:201–13. ~amar R. Measuring compliance Potter L, Oakley D, de Leon-Wong E, Can among oral contraceptive users. Fam Plann Perspect 1996;28:154–8. Josefsson A, Wir ehn A-B, Lindberg M, Foldemo A, Brynhildsen J. Continuation rates of oral hormonal contraceptives in a cohort of first-time users: a population-based registry study, Sweden 2005–2010. BMJ Open 2013;3: e003401. Sanders SA, Graham CA, Bass JL, Bancroft J. A prospective study of the effects of oral contraceptives on sexuality and well-being and their relationship to discontinuation. Contraception 2001;64:51–8. Ott MA, Shew ML, Ofner S, Tu W, Fortenberry JD. The influence of hormonal contraception on mood and sexual interest among adolescents. Arch Sex Behav 2008;37:605–13. Hall KS, O'Connell White K, Rickert VI, Reame N, Westhoff C. Influence of depressed mood and psychological stress symptoms on perceived oral contraceptive side effects and discontinuation in young minority women. Contraception 2012;86:518–25. Ernst U, Baumgartner L, Bauer U, Janssen G. Improvement of quality of life in women using a low-dose desogestrel-containing contraceptive: results of an observational clinical evaluation. Eur J Contracept Reprod Health Care 2002; 7:238–43. Kelly S, Davies E, Fearns S, McKinnon C, Carter R, Gerlinger C, et al. Effects of oral contraceptives containing ethinylestradiol with either drospirenone or levonorgestrel on various parameters associated with well-being in healthy women. Clin Drug Investig 2010;30:325–36. Poromaa IS, Segebladh B. Adverse mood symptoms with oral contraceptives. Acta Obstet Gynecol Scand 2012;91:420–7. Graham CA, Sherwin BB. The relationship between mood and sexuality in women using oral contraceptive as a treatment for premenstrual symptoms. Psychoneuroendocrinology 1993;18:273–81. Graham CA, Ramos R, Bancroft J, Maglaya C, Farley TM. The effects of steroidal contraceptives on the we1l-being and sexuality of women: a double-blind, placebo-controlled, two-centre study of combined and progestogen-only methods. Contraception 1995;52:363–9. O’Connell K, Davis AR, Kerns J. Oral contraceptives: side effects and depression in adolescent girls. Contraception 2007;75:299–304. €m J, Fredrikson M, et al. Oral Gingnell M, Engman J, Frick A, Moby L, Wikstro contraceptive use changes brain activity and mood in women with previous negative affect on the pill–—a double-blinded, placebo-controlled randomized trial of a levonorgestrel-containing combined oral contraceptive. Psychoneuroendocrinology 2013;38:1133–44. Cullberg J. Mood changes and menstrual symptoms with different gestagen/estrogen combinations: a double blind comparison with a placebo. Acta Psychiatr Scand Suppl 1972;236:1–86.
19.
20.
21. 22.
23. 24. 25. 26.
27. 28. 29.
30.
31.
32.
33.
34.
Skovlund CW, Mørch LS, Kessing LV, Lidegaard Ø. Association of hormonal contraception with depression. JAMA Psychiatry 2016;73:1154–62. Chassany O, Dimen€as E, Dubois D, Wu A, Dupuy H. The Psychological General Well-Being Index (PGWBI) user manual. Lyon, France: MAPI Research Institute; 2004. Beck AT, Steer RA, Ball R, Ranieri WF. Comparison of Beck Depression Inventories -IA and -II in psychiatric outpatients. J Pers Assess 1996;67:588–97. Lundgren-Nilsson A, Jonsdottir IH, Ahlborg G Jr, Tennant A. Construct validity of the psychological general well-being index (PGWBI) in a sample of patients undergoing treatment for stress-related exhaustion: a Rasch analysis. Health Qual Life Outcomes 2013;11:2. Ke Y, Bertin J, Gonthier R, Simard JN, Labrie F. A sensitive, simple and robust LC–MS/MS method for the simultaneous quantification of seven androgenand estrogen-related steroids in postmenopausal serum. J Steroid Biochem Mol Biol 2014;144:523–34. €derg €m T, Shanbag V, Carstensen H. Calculation of free and So ard R, B€ackstro bound fractions of testosterone and estradiol-17 beta to human plasma proteins at body temperature. J Steroid Biochem 1982;16:801–10. Cohen J. Statistical power analysis for the behavioral sciences. 2nd ed. Mahwah: NJ: Lawrence Erlbaum Associates; 1988. Shifren JL, Braunstein GD, Simon JA, Casson PR, Buster JE, Redmond GP, et al. Transdermal testosterone treatment in women with impaired sexual function after oophorectomy. N Engl J Med 2000;343:682–8. Somboonporn W, Davis S, Seif MW, Bell R. Testosterone for peri and postmenopausal women. Cochrane Database Syst Rev 2005:CD004509. Compagnone NA, Mellon SH. Neurosteroids: biosynthesis and function of these novel neuromodulators. Front Neuroendocrinol 2000;21:1–56. Simoncini T, Genazzani AR. Non-genomic actions of sex steroid hormones. Eur J Endocrinol 2003;148:281–92. Turkmen S, B€ackstrom T, Wahlstrom G, Andreen L, Johansson IM. Tolerance to allopregnanolone with focus on the GABA-A receptor. Br J Pharmacol 2011;162:311–27. Rapkin AJ, Akopians AL. Pathophysiology of premenstrual syndrome and premenstrual dysphoric disorder. Menopause Int 2012;18:52–9. €m T, Bixo M, Johansson M, Nyberg S, Ossewaarde L, Ragagnin G, et al. B€ackstro Allopregnanolone and mood disorders. Prog Neurobiol 2014;113:88–94. Joffee H, Cohen LS, Harlow BL. Impact of oral contraceptive pill use on premenstrual mood: predictors of improvement and deterioration. Am J Obstet Gynecol 2003;189:1523–30. Lidegaard O, Hougaard Nielsen L, Wessel Skovlund C, Lokkegaard E. Venous thromobosis in users of non-oral hormonal contraception: followup study, Denmark 2001–10. BMJ 2012;344:e2990. Kurshan N, Epperson CN. Oral contraceptives and mood in women with and without premenstrual dysphoria: a theoretical model. Arch Womens Ment Health 2006;9:1–14. Caruso S, Sareri MI, Agnello C, Romano M, Lo Presti L, Malandrino C, et al. Conventional vs. extended-cycle oral contraceptives on the quality of sexual life: comparison between two regimens containing 3 mg drospirenone and 20 mg ethinyl estradiol. J Sex Med 2011;8:1478–85. Zethraeus N, Dreber A, Ranehill E, Blomberg L, Labrie F, von Schoultz B, et al. Combined oral contraceptives and sexual function in women–a doubleblind, randomized, placebo-controlled Trial. J Clin Endocrinol Metab 2016;101:4046–53. Lopez LM, Kaptein AA, Helmerhorst FM. Oral contraceptives containing drospirenone for premenstrual syndrome. Cochrane Database Syst Rev 2012: CD006586.
VOL. - NO. - / - 2017
A first-choice combined oral contraceptive influences general well-being in healthy women: a double-blind, randomized, placebo-controlled trial Niklas Zethraeus, Ph.D.,a Anna Dreber, Ph.D.,b Eva Ranehill, Ph.D.,c,d Liselott Blomberg, B.Sc.,e Fernand Labrie, M.D., Ph.D.,f Bo von Schoultz, M.D., Ph.D.,d,e Magnus Johannesson, Ph.D.,b n Hirschberg, M.D., Ph.D.e,g and Angelica Linde a Department of Learning, Informatics, Management and Ethics, Karolinska Institutet, Stockholm, Sweden; b Department of Economics, Stockholm School of Economics, Stockholm, Sweden; c Department of Economics, University of Zurich, Zurich, Switzerland; d Department of Economics, University of Gothenburg, Gothenburg, Sweden;e Department of Obstetrics and Gynecology, Karolinska University Hospital, Stockholm, Sweden; f Laval University, Quebec City, Montreal, Canada; and g Department of Women's and Children's Health, Karolinska Institutet, Stockholm, Sweden
Objective: To determine whether there is a causal effect of oral contraceptive (OC) treatment on general well-being and depressed mood in healthy women. Design: Double-blind, randomized, and placebo-controlled trial. Setting: University hospital. Patient(s): Three hundred and forty healthy women aged 18–35 years randomized to treatment, of whom 332 completed the data collection at follow-up evaluation. Intervention(s): A combined OC (150 mg levonorgestrel and 30 mg ethinylestradiol) or placebo for 3 months of treatment. Main Outcome Measure(s): Primary outcome measures: global score of Psychological General Well-Being Index (PGWBI) and the Beck Depression Inventory (BDI); secondary outcome measures: six separate dimensions of the PGWBI. Result(s): The OC treatment statistically significantly decreased general well-being compared with placebo 4.12 (95% CI, 7.18 to 1.06). Furthermore, OC decreased the following PGWBI dimensions compared with placebo: positive well-being 3.90 (95% CI, 7.78 to 0.01), self-control 6.63 (95% CI, 11.20 to 2.06), and vitality 6.84 (95% CI, 10.80 to 2.88). The effect of OC on depressive symptoms and on the PGWBI dimension depressed mood were not statistically significant. Conclusion(s): This study demonstrates a statistically significant reduction in general well-being by a first-choice OC in comparison with placebo in healthy women. We found no statistically significant effects on depressive symptoms. A reduction in general well-being should be of clinical importance. (Fertil SterilÒ 2017;-:-–-. Ó2017 by American Society for Reproductive Medicine.) Key Words: Depression, oral contraceptives, quality of life, randomized clinical trial, well-being Discuss: You can discuss this article with its authors and with other ASRM members at https://www.fertstertdialog.com/users/ 16110-fertility-and-sterility/posts/15214-23476
Received November 29, 2016; revised January 28, 2017; accepted February 24, 2017. N.Z. has nothing to disclose. A.D. has nothing to disclose. E.R. has nothing to disclose. L.B. has nothing to disclose. F.L. has received personal fees from EndoCeutics Inc. B.V.S. has nothing to disclose. M.J. has nothing to disclose. A.L.H. has nothing to disclose. Supported by research grants from Jan Wallander and Tom Hedelius Foundation (P2010–0133:1, P2012–0002:1, and P2013–0156:1), Knut and Alice Wallenberg Foundation (Wallenberg Academy Fellows grant to A. Dreber), Swedish Council for Working Life and Social Research (2006–1623), the Swiss National Science Foundation (100010–149451), the Swedish Research Council (20324), Karolinska Institutet, and the regional agreement on medical training and clinical research (ALF) between Stockholm County Council and Karolinska Institutet (20130313). The funders of the study had no role in study design, data collection, data analysis, data interpretation, or writing of the report. All authors had full access to all the data in the study and were responsible for the decision to submit for publication. N.Z., A.D., and E.R. should be considered similar in author order. n Hirschberg, M.D., Ph.D., Department of Obstetrics and Gynecology, Karolinska University Hospital, Stockholm SE-171 76, Reprint requests: Angelica Linde Sweden (E-mail: [email protected]). Fertility and Sterility® Vol. -, No. -, - 2017 0015-0282/$36.00 Copyright ©2017 American Society for Reproductive Medicine, Published by Elsevier Inc. http://dx.doi.org/10.1016/j.fertnstert.2017.02.120 VOL. - NO. - / - 2017
1
ORIGINAL ARTICLE: CONTRACEPTION
T
he introduction of the oral contraceptive pill (OC) more than 50 years ago has had a tremendous impact on women's reproductive health and position in society. Today, it is estimated that over 100 million women worldwide are current users (1). Although OC pills are recognized as highly effective and generally safe, there is still remarkably poor adherence to this contraceptive method. In fact, numerous surveys show a high discontinuation rate and irregular use in as many as 60% of women (1–3). Poor adherence and contraceptive failure have been related to certain side effects such as negative mood changes and depressive symptoms (4–6). Some 4% to 10% of OC users report depressive symptoms (7–9). However, very few placebo-controlled studies have addressed this issue, and they have been performed in women with premenstrual symptoms (10), sterilized women (11), women with dysmenorrhea (12), and in those with previous depressive symptoms using OCs (13). Most of these relatively small studies have not found any statistically significant effect of OC treatment on depressive symptoms (10–12). However, one placebocontrolled study in healthy women demonstrated a statistically significant impairment of mood symptoms by OC use (14); another study showed worsening of depressive symptoms in women with previous experience of negative mood symptoms while taking OCs (13). Recently, a large cohort study also demonstrated an association between hormonal contraception and subsequent use of antidepressants (15). The daily intake of OCs and a positive attitude of the user are a necessary prerequisites for efficacious oral contraception. Irregular intake might reflect a negative influence on the individual woman's well-being in a more general sense than specific depressive symptoms. Apart from depressed mood, well-being also implies dimensions like vitality, selfcontrol, and perception of general health. There is very little information about causal effects of OC use on general wellbeing in healthy women. We found only two small placebocontrolled studies, both with inconclusive results (10, 11). We have filled a knowledge gap regarding a possible causal effect of OC treatment on general well-being and depressive symptoms in healthy women. We performed a double-blind, randomized, placebo-controlled trial on a combined OC in 340 young, healthy women. Our hypothesis was that OCs decrease general well-being and increase depressive symptoms (7–9, 13–15).
MATERIALS AND METHODS Study Design The study was a single center, randomized, double-blind, placebo-controlled, and parallel study conducted at the Karolinska University Hospital, Stockholm, Sweden. The study was performed in accordance with good clinical practice and the World Medical Association Declaration of Helsinki's ethical principles for medical research involving human subjects. The protocol and the informed consent form were approved by the local ethics committee in Stockholm (2010/1075-31/ 1) and the Swedish Medical Products Agency (151:2010/ 44213). The trial is registered at the EudraCT registry, registration number EudraCT 2010-020824-23. 2
Study Population The sample comprised women willing to start using OCs. The inclusion criteria were 18 to 35 years of age, body mass index (BMI) 19–30, a regular menstrual cycle (25–33 days), using nonhormonal contraceptive at the study start, and fluency in the Swedish language. The exclusion criteria were smoking, hypertension, risk factors for thromboembolism and cardiovascular disease, severe migraine, diabetes mellitus, liver disease, pancreatitis, history of hormone-dependent cancer, undiagnosed bleeding, and pregnancy. Intake of any sex steroid hormones during the last 6 weeks before beginning the study was not allowed. Those fulfilling all the inclusion criteria and none of the exclusion criteria were included in the study. All women were instructed to use nonhormonal contraceptive methods during the study. Free condoms were distributed.
Randomization Participants were randomly assigned to treatment with a combined monophasic pill (Neovletta; Bayer Schering Pharma), containing 150 mg of levonorgestrel and 30 mg of ethinylestradiol, or placebo during a period of 3 months. Randomization was provided by a contract organization (APL, Pharma Specials). A balanced block randomization [1:1] was performed with 10 participants in each block of fixed size. The study investigators, research coordinators, and the participating women were blinded to the treatment allocation. Participants took one capsule every day for 21 days, followed by a break during 7 days. Each 28-day period defined one treatment cycle.
Procedures During the baseline visit, scheduled for the early follicular phase of the menstrual cycle (cycle days 1–7), a general health control was performed that included weight, height, and blood pressure measurements. Participants also provided data on education and marital status. At baseline, a blood sample (20 mL) was collected for analyses of hormones and binding proteins. After the blood sample collection, participants filled out two surveys on general well-being and depressive symptoms: the Psychological General Well-Being Index (PGWBI) (16) and the Beck Depression Inventory (BDI) (17). The final data collection was performed at the end of the third treatment cycle. Participants underwent the same procedures as at baseline: weight and blood pressure measurements, blood sampling for hormone determination, and assessment of general well-being and depressive symptoms. All the surveys were computerized and were completed by the participants at the study center between 08:00 and 18:00 the same day.
Outcomes Our primary outcome measure of general well-being was the normalized global score of the PGWBI (16), and our secondary outcome measures of general well-being were the normalized scores on each PGWBI dimension. The PGWBI measured VOL. - NO. - / - 2017
Fertility and Sterility® general well-being during the last month and consists of 22 items, where each item has six response choices and the value of each item ranges from 0 to 5. This scale has had satisfactory internal construct validity when tested with modern psychometric techniques (18). The PGWBI consists of the following six dimensions: anxiety (included five items), depressed mood (three items), positive well-being (4 items), self-control (3 items), general health (3 items), and vitality (4 items). The global PGWBI score was computed as the sum of scores on each item (that is, each of the 22 items was given the same weight). The global score of PGWBI varies between 0 (poor quality of life) and 110 (good quality of life). A score of %72 in PGWBI indicates moderate to severe distress in general well-being. The score on each PGWBI dimension (anxiety, depressed mood, positive well-being, self-control, general health, and vitality) was defined as a secondary outcome measure. The total score for the different dimensions was anxiety (0–25), depressed mood (0–15), positive wellbeing (0–20), self-control (0–15), general health (0–15), and vitality (0–20). To facilitate comparison across studies and for comparisons with other quality-of-life instruments, the total score on each dimension and the global score of PGWBI was normalized to 0–100. The global score of the BDI (17) was our primary outcome measure of depressive symptoms. The BDI consists of 21 items, where each item has four response choices, and the value of each item ranges from 0 to 3. The BDI measured the following 21 depressive symptoms and attitudes during the last week: sadness, pessimism/discouragement, sense of failure, dissatisfaction, guilt, expectation of punishment, self-dislike, selfaccusation, suicidal ideation, crying, irritability, social withdrawal, indecisiveness, body-image distortion, work retardation, insomnia, fatigability, anorexia, weight loss, somatic preoccupation, and loss of libido. A global score of BDI is computed as the sum of scores on each item, and ranges from 0–63, where a higher value reflects more severe depressive symptoms/attitudes. A score R20 in the BDI indicates moderate to severe depressive symptoms.
SHBG, and a fixed albumin concentration based on an equation system derived from the law of mass action (20).
Statistical Analysis Before completing the data collection and before breaking the randomization codes we posted a predefined analysis protocol at the Open Science Framework Web site (freely available at https://osf.io/he8nb/) on August 21, 2015. The main analysis was performed as an intention-to-treat analysis (ITT) and included all women who completed the data collection. To test whether there was a statistically significant difference in the average value of an outcome measure between the OC group and the placebo group, we used an independent samples t-test (not assuming equal variances). The tests were based on the difference in the value of the outcome measure between the final data collection and the baseline data collection for each participant, and all statistical tests were two-sided (alpha level, 0.05). We specified two subgroup analyses in the protocol for two subgroups of moderate to severe distress in general well-being and depressive symptoms. These two subgroups were defined as those who had scores of %72 in PGWBI (moderate to severe distress) before normalization of the score and those who had scores R20 in BDI (severity level: moderate to severe). In exploratory analyses, we also estimated the correlation between the pretreatment and posttreatment change in serum levels of total and free testosterone and the change in general well-being and depressive symptoms, using the Pearson and the Spearman correlation coefficients. The sample size of 340 gives us a statistical power of 80% to find a statistically significant difference at the 5% level in general well-being (PGWBI) of 4.3 units measured on the PGWBI normalized scale between 0 and 100, and 2.1 units in depressive symptoms measured on the BDI scale between 0 and 63. In the Supplemental Appendix we provide a complete methods section with details about the study population, study design, treatment, general well-being and depressive symptom measurements, safety assessments, hormone measurements, and statistical analysis.
Safety Assessments Adverse events and severity were assessed at each scheduled contact and whenever there was an event. The women were reminded to use nonhormonal contraception. At each visit, vital signs including blood pressure and weight were recorded, and a pregnancy test was performed.
Hormone Measurements Quantification of testosterone and estradiol in serum was performed by the validated, highly sensitive and specific liquid chromatography with tandem mass spectrometry (LC-MS/ MS) method performed at Endoceutics (Quebec, Canada) (19). Serum concentrations of follicle-stimulating hormone (FSH) and luteinizing hormone (LH) were determined by fluorescence immunoassays (Auto Delfia; Perkin Elmer), and sex-hormonebinding globulin (SHBG) by chemiluminescent enzyme immunometric assay (Roche Diagnostics). Concentrations of free testosterone were calculated from values of total testosterone, VOL. - NO. - / - 2017
RESULTS About 1,000 women, were initially screened by phone, and 347 women were assessed for eligibility at a screening visit. Three hundred and forty women were included in the study between February 3, 2012, and August 24, 2015, and were randomly assigned to either OC or placebo (169 in the OC group, 171 in the placebo group). Seven participants were lost to follow-up evaluation, and one participant did not complete the data collection due to an administrative error (Fig. 1). All the 332 women who completed the data collection at follow-up evaluation were included in our main analysis, including the three women who discontinued treatment but completed the data collection. There were no statistically significant differences in any of the baseline characteristics between the two treatment groups (P>.05) (Table 1). We hypothesized that OCs would reduce general wellbeing and increase depressive symptoms. Consistent with this hypothesis, there was a statistically significant difference in 3
ORIGINAL ARTICLE: CONTRACEPTION
FIGURE 1
340 Underwent randomization
169* Were assigned to and received oral contraceptives (150 μg levonorgestrel and 30 μg ethinylestradiol)
4 Discontinued drug and were lost to follow-up - 1 anxiety - 1 increased appetite - 1 depression - 1 panic disorder 1 Continued drug but did not complete data collection due to an administrative error
2 Discontinued drug but completed data collection† - 1 mood changes - 1 lost the pills
162 Continued drug and completed data collection†
171 Were assigned to and received placebo
3 Discontinued placebo and were lost to follow-up - 1 forgot pills when travelling - 1 depression - 1 unknown
1 Discontinued placebo but completed data collection† - 1 nausea
167 Continued placebo and completed data collection†
Trial profile.* One participant initially randomized to the treatment group discontinued treatment due to an egg donation but was later randomized into the study again. This explains the uneven number of participants in the two groups. yAll participants that completed the data collection at follow-up were included in our main analysis (N¼164 in the oral contraceptive group and N¼168 in the placebo group). Zethraeus. Oral contraceptives and well-being. Fertil Steril 2017.
the change in the global score of PGWBI between the OC and the placebo groups of 4.12 (95% CI, 7.18 to 1.06; P¼ .0085) (Fig. 2A). We also found a statistically significant negative impact of OC on well-being in the hypothesized direction for the following three PGWBI dimensions (Fig. 2B): positive well-being 3.90 (95% CI, 7.78 to 0.01; P¼ .0492), self-control 6.63 (95% CI, 11.20 to 2.06; P¼ .0046), and vitality 6.84 (95% CI, 10.80 to 2.88; P¼ .0008). The effect sizes expressed as Cohen's D (21) were 0.29 for the global PGWBI score, 0.22 for reduced positive well-being, 0.34 for self-control, and 0.41 for the vitality dimension. These effect sizes are considered small to medium. For the other three PGWBI dimensions, anxiety 2.08 (95% CI, 6.15 to 1.99; P¼ .3155), depressed mood 3.67 (95% CI, 7.51 to 0.17; P¼ .0613), and general health 2.15 (95% CI, 5.24 to 0.94; P¼ .1723), we did not find a statistically significant difference between the OC and the placebo group (Fig. 2B). For the BDI, our primary outcome measure of depressive symptoms, the effect was not statistically significant 0.85 (95% CI, 0.66 to 2.36; P¼ .2664) (Fig. 3). 4
The effect size and statistical significance only changed marginally in our two robustness tests when we also included individuals who did not complete the data collection (n ¼ 340) or only included participants who did not discontinue treatment (n ¼ 329) (P¼ .0088 and .0073 for the PGWBI; .0502 and .0489 for PGWBI dimension ‘‘positive well-being’’; .0048 and .0047 for the PGWBI dimension ‘‘self-control’’; and finally .0008 and .0005 for the PGWBI dimension ‘‘vitality’’). Neither the BDI (P¼ .2684 and .2643), nor any of the other PGWBI domains (anxiety: P¼ .3168 and .3189; depressed mood: P¼ .0622 and .0583; and general health: P¼ .1745 and .1110) were statistically significant in any of the two robustness tests. The proportion of women with moderate to severe distress in general well-being (%72 on the PGWBI) and with moderate to severe depressive symptoms (R20 on the BDI) was comparable between the groups before treatment (Table 1). We found that 35% the OC and placebo groups had moderate to severe distress in general well-being at baseline, whereas 7% in the OC and placebo group had moderate to severe VOL. - NO. - / - 2017
Fertility and Sterility®
TABLE 1 Baseline characteristics of the intention-to-treat population. Characteristic
Placebo (n [ 171)
OC (n [ 169)
Age (y) 24.1 4.0 23.3 3.6 Educationa Primary school 5 (2.9) 6 (3.6) Secondary school 53 (31.2) 58 (34.7) University studies 64 (37.6) 56 (33.5) (ongoing) University degree 48 (28.2) 47 (28.1) In a stable relationshipa Yes 100 (58.8) 97 (57.7%) 22.6 2.8 22.5 3.1 BMI (kg/m2) Participants with moderate to severe distress in general well-being %72 in the PGWBI 60 (35.1) 59 (34.9) Participants with moderate to severe depressive symptoms R20 in the BDI 12 (7.0) 12 (7.1) 68.5 14.6 69.0 13.5 Score on PGWBIb b Scores on PGWBI dimensions Anxiety 63.5 18.2 62.9 18.5 Depressed Mood 79.3 16.3 78.8 15.2 Positive Well-Being 60.7 18.0 61.6 17.3 Self-Control 71.1 19.6 71.7 18.9 General Health 84.5 13.6 84.7 12.9 Vitality 60.2 17.1 63.0 15.8 c 7.3 7.3 7.8 6.9 Score on BDI Serum levels of blood sample testsd FSH (IE/L) 5.4 1.7 5.4 1.7 LH (IE/L) 4.9 2.4 4.8 2.2 Total T (pg/mL) 246.4 93.7 244.3 88.7 SGBG (nmol/L) 78.1 46.7 72.6 27.3 Free T (pg/mL) 3.2 1.3 3.2 1.3 Estradiol (pg/mL) 33.5 19.1 34.4 25.1 Note: Data presented as mean standard deviation or n (%). The body mass index (BMI) is the weight in kilograms divided by the square of the height in meters. There were no statistically significant differences between the two treatment groups (P>.05), as assessed with the use of the chi-square test for categorical data and the independent samples t-test (not assuming equal variances) for continuous data. BDI ¼ Beck Depression Inventory; BMI ¼ body mass index; FSH ¼ follicle-stimulating hormone; LH ¼ luteinizing hormone; OC ¼ oral contraceptive; PGWBI ¼ Psychological General Well-Being Index; SHBG ¼ sex hormone-binding globulin; T ¼testosterone. a There were three missing values on the education variable—two in the OC group and one in the placebo group—and two missing values on the marital status variable (one in the OC group and one in the placebo group). b The PGWBI (and each PGWBI dimension) ranges from 0 to 100, with higher values indicating greater general well-being. The PGWBI is defined as the average of the normalized score on each PGWBI item where each item is given the same weight. c The Beck Depression Inventory (BDI) scores range from 0 to 63, with higher scores indicating more severe depressive symptoms. d There were two missing values in the OC group for FSH, LH, SHBG, and free T and one missing value for total T and estradiol in the OC group. Zethraeus. Oral contraceptives and well-being. Fertil Steril 2017.
depressive symptoms. After treatment, 44% in the OC group and 38% in the placebo group had moderate to severe distress in general well-being, whereas 7% in the OC group and 7% in the placebo group had moderate to severe depressive symptoms. There was no statistically significant treatment effect for the primary outcome measures or the secondary outcome measures in these subgroups, but it should be noted that the study was not powered for comparisons within subgroups. As expected, the serum levels of FSH, LH, and total and free testosterone decreased after 3 months of active treatment, and the change in hormone levels was statistically significantly different between the OC and the placebo groups, supporting that the participants were compliant with the study treatment (independent samples t-test [not assuming equal variances], P< .001) (Supplemental Table 1). VOL. - NO. - / - 2017
In exploratory analyses, we tested whether the size of the pretreatment and posttreatment changes in serum levels of total and free testosterone were correlated with the change in general well-being and depressive symptoms. None of these correlations were statistically significant in the OC group (Pearson and Spearman correlations, P>.05, twosided), and only two correlations were statistically significant in the placebo group (Supplemental Tables 2 and 3). We detected adverse events in 21.0% of the women in the OC group and 12.0% in the placebo group (Supplemental Table 4). There were four serious adverse events in total, including two pregnancies and one ovarian cyst in the placebo group and one panic disorder in the OC group. The most common reported adverse events in the OC group were bleeding disturbances (n ¼ 14), anxiety and mood changes (n ¼ 12), acne (n ¼ 5), and appetite changes (n ¼ 3). Anxiety and mood changes (n ¼ 4) were the most common reported adverse events in the placebo group.
DISCUSSION Here we report results from the first large randomized, placebo-controlled trial of a combined OC on general wellbeing and depressive symptoms in healthy women. We found a statistically significant reduction in general well-being in women who used a levonorgestrel-containing OC for 3 months compared with placebo, as measured with the established PGWBI instrument. There were also negative effects on the specific PGWBI dimensions of positive well-being, selfcontrol, and vitality. Thus, the results were consistent with our hypothesis that in some women OC use could cause a decrease in general well-being. The difference in the primary outcome measure PGWBI between the OC and the placebo group corresponds to a 6.0% reduction of the baseline value. The corresponding reductions in the three significant PGWBI dimensions (positive well-being, self-control, and vitality) were 6.4%, 9.3%, and 11.1%, respectively. These small to medium reductions, in terms of Cohen's D terminology (21), in well-being are likely to be of clinical importance for individual women and may be one explanation for the high discontinuation rate. For the other primary outcome measure, depressive symptoms as assessed by the BDI, we found no statistically significant negative effect. Thus, our hypothesis of increased depressive symptoms could not be confirmed. Moreover, the PGWBI dimension of depressed mood was not statistically significantly different from the controls. The nonstatistically-significant effect on depressive symptoms is in line with most previous placebo-controlled studies (10–12). However, higher scores of depressed mood with OC treatment in comparison with placebo have been reported in two previous placebo-controlled studies (13, 14). Furthermore, an association between hormonal contraception and a first diagnosis of depression, especially among adolescents, was recently suggested in a large cohort study (15). The mechanism behind impaired well-being by OC use in healthy women is not known. We found an increase in SHBG and consequently a decrease in total and free testosterone 5
ORIGINAL ARTICLE: CONTRACEPTION
FIGURE 2
Mean changes from baseline to end of follow-up in scores for the Psychological General Well-Being Index (PGWBI). (A) Results for the global score of PGWBI (the average of the score on each item). (B) Results for each PGWBI dimension. The global score of PGWBI and the score on each PGWBI dimension range from 0 to 100, with higher scores indicating increased well-being. Error bars denote 95% confidence intervals. Zethraeus. Oral contraceptives and well-being. Fertil Steril 2017.
accompanying OC treatment. Some studies have demonstrated positive effects of testosterone treatment on general well-being in postmenopausal women (22, 23). Consequently, a reduction in free testosterone by OC use might cause reduced general well-being. However, we found no statistically significant correlations in the OC group between the change in serum levels of total and free testosterone and changes in general well-being or depressive symptoms. Another possible mechanism whereby OCs might affect general well-being is a direct negative effect of the progestin component on the brain. Sex hormones can be classified as neurosteroids because they can be synthesized in the central and peripheral nervous system (24). They bind to their specific nuclear receptors and act via genomic effects but could also influence brain function via more rapid nongenomic mechanisms and indirectly through neurotransmitter systems, including the serotonin and gamma-aminobutyric acid 6
(GABA) systems (25). Estrogen has been shown to enhance brain excitability, whereas progesterone seems to have an opposite and inhibitory influence (26). Progesterone and several progesterone metabolites exert a sedative effect on the brain. Some metabolites (e.g., allopregnanolone) could even be regarded as anesthetic agents (26). Clearly, such influence could affect the sense of vitality and reduce general wellbeing. Cyclic negative mood symptoms such as irritability, loss of impulse control, depression, and fatigue have been related to an increase of progestogenic neurosteroids during the luteal phase of the menstrual cycle (27, 28). Different progestins and combined OCs have been reported to provoke similar symptoms (29). Although we did not find statistically significant effects on depressive symptoms, it is possible that a direct progestin-induced central nervous system effect could be the underlying mechanism behind reduced well-being, self-control, and vitality in the OC group. VOL. - NO. - / - 2017
Fertility and Sterility®
FIGURE 3
Mean changes from baseline to end of follow-up in scores for the Beck Depression Inventory (BDI). The BDI ranges from 0 to 63, with higher scores indicating more depressive symptoms—that is, a positive change reflects an increase in depressive symptoms. Error bars denote 95% confidence intervals. Zethraeus. Oral contraceptives and well-being. Fertil Steril 2017.
Combined OCs are more or less equally effective as contraceptive agents, but they may differ in other aspects such as risk profile, side effects, and compliance. Most combined OCs contain the synthetic ethinylestradiol as an estrogen component, whereas a few recently developed contain estradiol. In contrast, there are several different progestins with varying degrees of androgenic activity combined with the estrogen component. Androgenic progestins in combination with ethinylestradiol (second-generation OCs) seem to be associated with a lower risk of venous thromboembolism and are therefore recommended as the first choice for contraception in some countries (30). In the present study, we used a levonorgestrel-containing OC, which is the drug most commonly prescribed in Sweden. On the other hand, negative side effects such as mood changes, acne, increased appetite, weight gain, and sexuality may be more frequent with androgenic OCs compared with antiandrogenic OCs (third- and fourth-generation OCs) (31-33). Antiandrogenic OCs may even have positive effects on mood. In placebo-controlled trials, combined OCs containing drospirenone (antiandrogenic fourth-generation OC) have demonstrated beneficial effects in women with premenstrual syndrome (34). Also, a shorter hormone-free interval by the 24/4 regimen may contribute to the beneficial effect because adverse mood symptoms are most pronounced during the pillfree interval (9). Thus, our findings may not be valid for all OCs. Additional studies on antiandrogenic OCs are needed. If the results from alternative OCs show less or no adverse effect on general well-being, this may be reason enough to reconsider recommendations of first-choice combined contraception. The strengths of our present study are the large number of participants and the double-blind placebo-controlled design. VOL. - NO. - / - 2017
However, one limitation was the short study period because it is well-known that several adverse effects vanish over time. Still, mood changes usually do not disappear, and a further deterioration is also possible. Furthermore, positive and negative expectations by the women under treatment may have influenced the results. The use of OCs will change the amount of and cyclicity of menstrual bleeding and thus patient blinding in reality may not have been complete. However, it is also known that women who usually have regular cycles can become anovulatory during a test situation. Even if questions on menstrual cyclicity were avoided, women may have suspected that they were given active or inactive treatment. Such bias, however, could work in both directions. Additional strengths of the study are the use of well-established and validated instruments for assessment of general well-being (PGWBI) and depressed mood (BDI), and the highly specific, validated, and accurate method for testosterone analysis.
CONCLUSION Our present study demonstrates a statistically significant reduction in general well-being with the use of a popular levonorgestrel-containing OC for 3 months in comparison with placebo in healthy women. Furthermore, the PGWBI dimensions of positive well-being, self-control, and vitality were reduced. However, we found no statistically significant effects on depressive symptoms in this group of healthy women. Even a modest reduction in general well-being during OC intake should be of clinical importance and may explain the high discontinuation rate and irregular use of this contraceptive method. This possible side-effect thus has to be recognized and discussed during contraceptive counseling. It is vitally important to evaluate alternative OC 7
ORIGINAL ARTICLE: CONTRACEPTION combinations in large placebo-controlled studies to identify those with the most beneficial profile. Acknowledgments: The authors thank Elisabeth Berg and Mesfin Tessma for valuable comments on the analysis protocol; Berit Legerstam, Siv R€ odin Andersson, and Agneta Berge for research assistance; and Johan Almenberg for proofreading the manuscript.
15. 16.
17.
18.
REFERENCES 1. 2. 3.
4.
5.
6.
7.
8.
9. 10.
11.
12. 13.
14.
8
Brynhildsen J. Combined hormonal contraceptives: prescribing patterns, compliance, and benefits versus risks. Ther Adv Drug Saf 2014;5:201–13. ~amar R. Measuring compliance Potter L, Oakley D, de Leon-Wong E, Can among oral contraceptive users. Fam Plann Perspect 1996;28:154–8. Josefsson A, Wir ehn A-B, Lindberg M, Foldemo A, Brynhildsen J. Continuation rates of oral hormonal contraceptives in a cohort of first-time users: a population-based registry study, Sweden 2005–2010. BMJ Open 2013;3: e003401. Sanders SA, Graham CA, Bass JL, Bancroft J. A prospective study of the effects of oral contraceptives on sexuality and well-being and their relationship to discontinuation. Contraception 2001;64:51–8. Ott MA, Shew ML, Ofner S, Tu W, Fortenberry JD. The influence of hormonal contraception on mood and sexual interest among adolescents. Arch Sex Behav 2008;37:605–13. Hall KS, O'Connell White K, Rickert VI, Reame N, Westhoff C. Influence of depressed mood and psychological stress symptoms on perceived oral contraceptive side effects and discontinuation in young minority women. Contraception 2012;86:518–25. Ernst U, Baumgartner L, Bauer U, Janssen G. Improvement of quality of life in women using a low-dose desogestrel-containing contraceptive: results of an observational clinical evaluation. Eur J Contracept Reprod Health Care 2002; 7:238–43. Kelly S, Davies E, Fearns S, McKinnon C, Carter R, Gerlinger C, et al. Effects of oral contraceptives containing ethinylestradiol with either drospirenone or levonorgestrel on various parameters associated with well-being in healthy women. Clin Drug Investig 2010;30:325–36. Poromaa IS, Segebladh B. Adverse mood symptoms with oral contraceptives. Acta Obstet Gynecol Scand 2012;91:420–7. Graham CA, Sherwin BB. The relationship between mood and sexuality in women using oral contraceptive as a treatment for premenstrual symptoms. Psychoneuroendocrinology 1993;18:273–81. Graham CA, Ramos R, Bancroft J, Maglaya C, Farley TM. The effects of steroidal contraceptives on the we1l-being and sexuality of women: a double-blind, placebo-controlled, two-centre study of combined and progestogen-only methods. Contraception 1995;52:363–9. O’Connell K, Davis AR, Kerns J. Oral contraceptives: side effects and depression in adolescent girls. Contraception 2007;75:299–304. €m J, Fredrikson M, et al. Oral Gingnell M, Engman J, Frick A, Moby L, Wikstro contraceptive use changes brain activity and mood in women with previous negative affect on the pill–—a double-blinded, placebo-controlled randomized trial of a levonorgestrel-containing combined oral contraceptive. Psychoneuroendocrinology 2013;38:1133–44. Cullberg J. Mood changes and menstrual symptoms with different gestagen/estrogen combinations: a double blind comparison with a placebo. Acta Psychiatr Scand Suppl 1972;236:1–86.
19.
20.
21. 22.
23. 24. 25. 26.
27. 28. 29.
30.
31.
32.
33.
34.
Skovlund CW, Mørch LS, Kessing LV, Lidegaard Ø. Association of hormonal contraception with depression. JAMA Psychiatry 2016;73:1154–62. Chassany O, Dimen€as E, Dubois D, Wu A, Dupuy H. The Psychological General Well-Being Index (PGWBI) user manual. Lyon, France: MAPI Research Institute; 2004. Beck AT, Steer RA, Ball R, Ranieri WF. Comparison of Beck Depression Inventories -IA and -II in psychiatric outpatients. J Pers Assess 1996;67:588–97. Lundgren-Nilsson A, Jonsdottir IH, Ahlborg G Jr, Tennant A. Construct validity of the psychological general well-being index (PGWBI) in a sample of patients undergoing treatment for stress-related exhaustion: a Rasch analysis. Health Qual Life Outcomes 2013;11:2. Ke Y, Bertin J, Gonthier R, Simard JN, Labrie F. A sensitive, simple and robust LC–MS/MS method for the simultaneous quantification of seven androgenand estrogen-related steroids in postmenopausal serum. J Steroid Biochem Mol Biol 2014;144:523–34. €derg €m T, Shanbag V, Carstensen H. Calculation of free and So ard R, B€ackstro bound fractions of testosterone and estradiol-17 beta to human plasma proteins at body temperature. J Steroid Biochem 1982;16:801–10. Cohen J. Statistical power analysis for the behavioral sciences. 2nd ed. Mahwah: NJ: Lawrence Erlbaum Associates; 1988. Shifren JL, Braunstein GD, Simon JA, Casson PR, Buster JE, Redmond GP, et al. Transdermal testosterone treatment in women with impaired sexual function after oophorectomy. N Engl J Med 2000;343:682–8. Somboonporn W, Davis S, Seif MW, Bell R. Testosterone for peri and postmenopausal women. Cochrane Database Syst Rev 2005:CD004509. Compagnone NA, Mellon SH. Neurosteroids: biosynthesis and function of these novel neuromodulators. Front Neuroendocrinol 2000;21:1–56. Simoncini T, Genazzani AR. Non-genomic actions of sex steroid hormones. Eur J Endocrinol 2003;148:281–92. Turkmen S, B€ackstrom T, Wahlstrom G, Andreen L, Johansson IM. Tolerance to allopregnanolone with focus on the GABA-A receptor. Br J Pharmacol 2011;162:311–27. Rapkin AJ, Akopians AL. Pathophysiology of premenstrual syndrome and premenstrual dysphoric disorder. Menopause Int 2012;18:52–9. €m T, Bixo M, Johansson M, Nyberg S, Ossewaarde L, Ragagnin G, et al. B€ackstro Allopregnanolone and mood disorders. Prog Neurobiol 2014;113:88–94. Joffee H, Cohen LS, Harlow BL. Impact of oral contraceptive pill use on premenstrual mood: predictors of improvement and deterioration. Am J Obstet Gynecol 2003;189:1523–30. Lidegaard O, Hougaard Nielsen L, Wessel Skovlund C, Lokkegaard E. Venous thromobosis in users of non-oral hormonal contraception: followup study, Denmark 2001–10. BMJ 2012;344:e2990. Kurshan N, Epperson CN. Oral contraceptives and mood in women with and without premenstrual dysphoria: a theoretical model. Arch Womens Ment Health 2006;9:1–14. Caruso S, Sareri MI, Agnello C, Romano M, Lo Presti L, Malandrino C, et al. Conventional vs. extended-cycle oral contraceptives on the quality of sexual life: comparison between two regimens containing 3 mg drospirenone and 20 mg ethinyl estradiol. J Sex Med 2011;8:1478–85. Zethraeus N, Dreber A, Ranehill E, Blomberg L, Labrie F, von Schoultz B, et al. Combined oral contraceptives and sexual function in women–a doubleblind, randomized, placebo-controlled Trial. J Clin Endocrinol Metab 2016;101:4046–53. Lopez LM, Kaptein AA, Helmerhorst FM. Oral contraceptives containing drospirenone for premenstrual syndrome. Cochrane Database Syst Rev 2012: CD006586.
VOL. - NO. - / - 2017